Wikipedia:Wikipedia Signpost/2010-09-27/In the news
Wikimedia moves into India, critical conference, Vandalism detection contest
India: Media speculation on country's future "Wiki-capital"; community newsletter released
Barry Newstead, the Wikimedia Foundation's Chief Global Development Officer, visited India last week to prepare the opening of the organization's first office outside the US, combined with the hiring of a "National Program Director" for India (see Signpost coverage: September 6 and August 30).
Newstead visited New Delhi, Mumbai and Bangalore. It was the first time a Wikipedia Meetup was ever organized in New Delhi and Mumbai. It brought together a mix of bloggers, Wikipedians and people generally interested in contributing to Wikipedia. Some 40 people attended the Mumbai and Bangalore Meetups and about 20 attended the Delhi meet, despite heavy monsoon rains disrupting the city. Along with WMF Trustee Bishakha Datta and advisory board member Achal Prabhala, Newstead addressed a press conference in Bangalore city.
The office is set to open in early 2011, and its location has not been announced yet. Yet various Indian media speculated which of the three cities might have the best chances. Bangalore, which is already the seat of the Indian Wikimedia chapter appears to be the front-runner according to The Hindu ("Wikipedia eyes Bangalore"), Times of India ("Wiki logs into India for its second office"), and The Economic Times ("Wikipedia aims to interact in top 10 Indian languages"). However, The Press Trust of India ("Wikipedia to open India office soon") said "Wikipedia" would launch the office "probably in Mumbai", while Daily News and Analysis ("Bangalore is in fray to become Wikipedia's India capital") mentioned Pune as a fourth possibility.
Time Out Mumbai quoted Prabhala on possible reasons that might still be "keeping Indians from becoming active contributors". The Economic Times described some of Newstead's personal background. Before becoming the Foundation's CGDO in June (see Signpost coverage), he had been a consultant at the non-profit Bridgespan Group, having quit a lucrative career at the Boston Consulting Group. Newstead explained: "I’ve always had a passion for education and knowledge. I enjoyed the training I got in consulting, but it was a personal decision to spend my career working on important social issues".
In related news, the first Indian Wikimedia Community Newsletter has been released last week. The publication contains a foreword by Bishdatta and Jimbo Wales (who already praised the Indian-language Wikipedias last month). The newsletter is 36 pages long in PDF format.
The newsletter reports the size of the Indian community: because of the large number of languages in India, there are about 20 different Wikipedias, as well as a similar number in the incubator. There is also significant Indian editing activity on the English Wikipedia (English is one of the official languages of India). There is a story on the activities of the Indian Wikipedia Chapter, and information on the activities of the various Indian-language WMF projects and the interactions between them. Some interesting tidbits:
- The first Indian-language wikis were in Oriya, Punjabi, Assamese, and Malayalam, all created in 2002.
- The Hindi Wikipedia is the largest of the Indian projects, with 56,000 articles.
- The Malayalam Wikipedia is one of the few languages (and the only one with a non-roman alphabet) to have its contents published as a CD.
Third "critical point of view" conference
On September 25 and 26, the third "CPOV" conference about Wikipedia took place in Leipzig, Germany. Unlike the previous installments that had been held in Bangalore in January and in Amsterdam in March (Signpost coverage), this one was held in a local language (Wikipedia:Ein kritischer Standpunkt).
As in the other conferences, most of the presentations were by academic researchers from the humanities. But this installment saw more participation by active Wikipedians, several of whom discussed with researchers in a "Wikipedia and Science" roundtable and on a "Wikipedia and Criticism" panel, and gave introductory Wikipedia workshops on the first day, supported by the local chapter, Wikimedia Germany.
The event generated coverage in several major German media. For example, TV news service Tageschau interviewed the director of the Leipzig university library (one of the conference sites) about Wikipedia, with one question being about the new Article feedback tool.
Vandalism detection competition
The "1st International Competition on Wikipedia Vandalism Detection" was held in Padua, Italy on 22–23 September 2010, to evaluate automated tools for detecting malicious edits on Wikipedia (workshop page), having received submissions from nine teams. It was based on a corpus of 32452 edits, sampled from one week of Wikipedia's recent changes, that had been classified by 753 human annotators recruited via Amazon Mechanical Turk, enabling the project to check the accuracy of the automatic vandalism detectors, for example by comparing true positives and false positives ("TP" and "FP" in the PDF file). One conclusion was that a combination of all 9 submitted detectors can perform better than any single detector. The combined detector has performance of FP=20% at a threshold level where TP=95% (right-hand chart on page 11), i.e. it would catch 19 out of 20 vandalism edits if it were allowed an error rate of wrongly classifying one in five legitimate edits as vandalism. A data point for higher detection levels is FP=35% at TP=98%. The paper also catalogued the different features that the detectors combine to assess the probability that a given edit is vandalism (one of the simplest characteristics, employed by seven of the nine contestants, is whether it was made by an anonymous editor).
Some other automatic vandalism detection tools were not part of the competition, among them vandal fighting bots like User:ClueBot that already operate on Wikipedia, the STiki tool developed at the University of Pennsylvania (whose author says that it would have finished second if it had participated) and a recently announced project at the University of Iowa.
Briefly
- How Wikipedians use sources: A paper that appeared in this month's issue of the journal "Information Research" ("Where does the information come from? Information source use patterns in Wikipedia") reported on a Web survey that had been answered by 108 Wikipedians in spring 2008. Based on the results, the author Isto Huvila distinguishes five groups of Wikipedians according to their use of references (e.g., whether they tend to find sources by online search or by visits to the library): "Investigators", "Surfers", the "Worldly-wise", "Scholars" and "Editors".
- Use of image donation analyzed: Five months after announcing a large scale upload of "no known copyright" images from its collection to Wikimedia Commons (Signpost coverage), the Brooklyn Museum has published a brief statistical analysis showing how often the images are used on different Wikipedias, and how often the various metadata fields in the image descriptions were changed or corrected by Wikimedians ("Where in the Wikiverse is the Brooklyn Museum?").
Discuss this story